Working Paper on Newcomb's Paradox
Newcomb's Paradox is Not a Paradox, by Matt Beller (U.S. SEC)
Newcomb’s Paradox, more appropriately called Newcomb’s Problem, is a scenario in which someone claims to be able to predict a decision that you will make, and you are then asked to make that decision. Since its popularization by Robert Nozick in 1969 , misconceptions regarding Newcomb’s Problem have been put forward and allowed to flourish, convoluting and distorting what should be a straightforward resolution. Although Nozick’s analysis of Newcomb’s Problem is thorough and eventually arrives at the proper praxeological conclusion, re-exploring it gives us an excellent opportunity to examine some of the most fundamental assumptions of praxeology and Austrian economics and, in the tradition of Henry Hazlitt, to explode the fallacies that surround the problem.
Comments (13)
Larry N. Martin
An Austrian on the SEC? Cool!
Published: April 19, 2007 1:52 PM
Kevin B.
Ludwig Von Mises writes:
Perhaps the Predictor has developed technology which it uses to communicate with itself in the future.
Published: April 19, 2007 3:14 PM
Brent
I'm more curious as to whether Justice demands that future technology be used to resurrect the dead in order to punish them for their crimes.
Published: April 19, 2007 5:57 PM
SciFi Reader
"If you take only the second box ... you are not someone I recognize as human." -- Isaac Asimov
How fitting then that I find Asimov's books to be unreadable uninspired trash.
Published: April 19, 2007 6:41 PM
Matt Beller
Kevin B,
Regarding the Predictor communicating with itself in the future, I don't think that's possible. While it is possible for the Predictor to write something down and read it again later (in the future) it is not possible for it to send a note to itself in the past. Without getting into it very deeply, I would rely on the notion of time as an irreversible flux to justify that argument.
Published: April 19, 2007 7:30 PM
Not Quite Austrian
The article illustrates well the most serious (and perhaps only) flaw I see commonly in the reasoning generally attributed as 'Austrian':
Axiomatic assumptions in your system of reasoning do not automatically make those assuptions true, no matter how well they have served you. And things that contradict your assuptions are not automatically false.
A "What if..." can _never_ be invalidated, because it is _by definition_ the assuption that some statement is true. And no matter how outlandish, the logical consquences can be explored. e.g. "What if 1+1 isn't 2" => "we can derive a contradiction between this assuption and the axioms of our mathematical system" => "the axioms of our mathematical system do not hold" => "our entire mathematical system falls apart and can not longer be used for reasoning under this 'what if' senerio".
And why do I keep seeing "man is not even capable of imagining ..." as part of various arguments. "man is not even capable of imagining X" => "X is not possible/true" or "man has a limited imagination" (or much more likely and my own personal favorite "the particular author has a limited imagination about the bounds of man's imagination").
Something being true for some cases doesn't not automatically make it true in all cases (e.g. compare the utility of Newtonian physics and the utility of human reason/action). The existence of irregularity doesn't not imply the non-existence of regularity.
A person just saying something does not automatically make it true, no matter who they are. Quoting Mises does not constitute a logical argument (unless of course you are quoting Mises's logical argument).
No wonder this is coming out of the SEC (I wonder if this guy is their mole to discredit the Austrian school by making it sound like a religion - "if you are a ... then you must ...").
And I would always choose only box B which would result in either me getting $1M or falsifying the predictor's claims (both of which are 'interesting' to me, while the possibilities resulting from the other choice are less so compared to the opportunity costs... but of coarse this is entirely subjective: I want the $1M, don't care about the $1K, and am open to the predictor's claims).
It's only a 'paradox' to conventional economics because they discredit subjective value. The problem and paradox are ultimately intended for an economics that will make the decision for you. The 'problem' is "What does economics say you should do in this situation", mistakening thinking that that is the domain of economics. If you could prove that the predictor's claims contradict the axioms of praxeology (which I'm not convinced they do), then all you can say is that praxeology has nothing more to say about this problem (in the way that Mises says praxeology has nothing to say about God) - it is not an economics problem.
btw, check out Asimov's 'zeroth law' to understand where he's coming from... actually I'll save you the trouble of looking it up, it puts humanity 'as a whole' further up the priority scale than individual humans.
Altogether (Asimov's laws of robotics):
(where the lower numbered ones override the higher, I've left out parts that state this)
0. A robot may not harm humanity, or, by inaction, allow humanity to come to harm.
1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey orders given it by human beings.
3. A robot must protect its own existence.
So in other words, these robots would be pictures of collectivist utilitarian obedience. Can you imagine the hell these robots would wrought! I think people on this site uniquely can more than most.
Which gives more credibility to my claim that this guy might be a SEC mole, because what 'Austrian' would knowningly quote someone who's ideal is the above robots. And since when is the Austrian school about intellectual obedience to the canon of Mises's words (which is the impression I get from reading the paper).
I went and reread the paper, and the Asimov ad hominem combined with the crap on the authors website (a flashing animated gif, an mp3 and nothing else), lead me to be very suspicious of this author (as if the SEC wasn't enough).
Published: April 19, 2007 10:22 PM
Kevin B.
Matt Beller, you said:
Mises may have been quite adept with some important subjects, but he never demonstrated significant expertise in quantum theory.
Published: April 20, 2007 12:17 AM
Matt Beller
Kevin B.,
I'll acknowledge that Mises was probably not an expert in quantum theory. I'll consider adding more support for my claim that time is irreversible, other than that Mises said so. While thinking about your comments, it occurred to me that it might make for an interesting paper to explain how, through the use of quantum theory and time travel, socialist central planners could solve the calculation problem by, after experiencing shortages or surpluses, sending messages to themselves in the past saying to produce less bread, more beef, less blankets, etc. They could also use their foreknowledge to avoid natural disasters, prevent all violence, and access new technological methods for economic production.
Not Quite Austrian,
Praxeological reasoning does indeed rely on a priori knowledge, i.e. that 1+1=2. The validity of "1+1=2" cannot be debated logically. If you consider that to be a problem with Austrian economics, I can't myself offer any other arguments, but you might explore the writings of authors much more knowledgable about the subject of praxeology than I am. Mises and Roderick Long come to mind.
It was indeed my goal to show that the Predictor's claims are incompatible with the axioms of praxeology. Because you remain unconvinced, I will attempt to bolster my argument in a future version of the article. If I do indeed succeed, my statement at the end that "if...you consider yourself an Austrian ['praxeologist' might be a more appropriate term], you must first reject the notion that the Predictor is capable of predicting your choice" would be logically valid.
Regarding Asimov, I'll point out that I did not use Asimov, his books, or his comment on Newcomb's Problem to justify any argument that I made. I presented his comment as an illustration of the problem in using subjective preferences to seemingly make a logical argument about something. If Asimov said "I prefer blueberry pie to cherry pie," he would not be presenting a logical argument on why blueberry pie is better than cherry. Additionally, I could not validly argue that Asimov was wrong, just as I did not argue that his justification for selecting two boxes was wrong (or right).
Thank you for your feedback, and I'm glad that you enjoyed the animation on my website.
Published: April 20, 2007 10:37 AM
Mort Utley
Matt,
What does a priori mean?
Published: April 20, 2007 12:43 PM
Kevin B.
Matt Beller,
That does sound like an interesting paper. It would be amusing if the socialist planners would likely continue to be inefficient, even with knowledge of the future.
Published: April 20, 2007 2:23 PM
lester
kevin- that's a good point. in fact, they probably did know certain of their things were likely to fail, but felt it was a good way to keep the beurocracy going for a time. I read an essay by one of the neo cons, wurmser ithink, saying in 97 that if we invade iraq it will be insane tribal / civil warfare, no question. but he supported war anyway than and in 03.
Published: April 21, 2007 9:37 AM
Travis Page
An SEC mole to discredit Austrian economics? That is quite a conspiracy theory.
I enjoyed the paper. As far as the socialists go, my guess is that any change in actions that socialists would use to correct their shortages would just cause other problems.
Published: April 26, 2007 3:59 AM
Ornette
I've been thinking about NC for a while, and today I posted this piece about it. It considers a broader version of the paradox that your paper -- the Predictor isn't always right, and we don't know whether such fanciful devices as time travel or clairvoyance are (or could be) involved. As a result, I don't reach as strong a conclusion as you do, but I do think it can be reduced to manageable form. I hope you find it interesting.
Published: May 17, 2007 8:28 AM